library(tidyverse)
library(kableExtra)
theme_set(theme_classic())

Outliers

df_outliers_control_group <- do.call(rbind, lapply(Sys.glob("../Control Group/Data/outliers/*.csv"), read_csv)) %>% 
  mutate(tournament = "control group")

df_outliers_complete_heterogeneity <- do.call(rbind, lapply(Sys.glob("../Complete Heterogeneity/Data/outliers/*.csv"), read_csv)) %>%
  mutate(tournament = "complete heterogeneity")

df_outliers_chaning_SDM <- do.call(rbind, lapply(Sys.glob("../Changing SDM/Data/outliers/*.csv"), read_csv)) %>%
  mutate(tournament = "changing SDM")

Combine outliers

outliers <- df_outliers_complete_heterogeneity %>%
  rbind(df_outliers_control_group) %>%
  rbind(df_outliers_chaning_SDM)

Cooperation rates

df_control_group <- do.call(cbind, lapply(Sys.glob("../Control Group/Data/*.csv"), read_csv)) %>%
  janitor::clean_names() %>%
  mutate(tournament = "control group",
         round = row_number()) %>%
  pivot_longer(seed_1024:seed_8, names_to = "Seed") 

df_complete_heterogeneity <- do.call(cbind, lapply(Sys.glob("../Complete Heterogeneity/Data/*.csv"), read_csv)) %>%
  janitor::clean_names() %>%
  mutate(tournament = "complete heterogeneity",
         round = row_number()) %>%
  pivot_longer(seed_1024:seed_8, names_to = "Seed")

df_changing_SDM <- do.call(cbind, lapply(Sys.glob("../Changing SDM/Data/*.csv"), read_csv)) %>%
  janitor::clean_names() %>%
  mutate(tournament = "changing SDM",
         round = row_number()) %>%
  pivot_longer(seed_1024:seed_8, names_to = "Seed")

Combine cooperation rates

coop_rates <- df_control_group %>%
  rbind(df_complete_heterogeneity) %>%
  rbind(df_changing_SDM) %>%
  mutate(Seed = str_remove(Seed, "seed_"))

Geom lines and geom points are not useful.

coop_rates %>%
  mutate(seed = as.factor(Seed)) %>%
  ggplot() +
  geom_point(aes(round, value, color = seed)) +
  facet_wrap(~tournament) +
  labs(title = "Average cooperation ratio per seed per tournament type.",
       x = "coop ratio",
       y = "round") +
  xlim(1000,9000)

coop_rates %>%
  mutate(seed = as.factor(Seed)) %>%
  ggplot() +
  geom_line(aes(round, value, color = seed)) +
  facet_wrap(~tournament) +
  labs(title = "Average cooperation ratio per seed per tournament type.",
       x = "coop ratio",
       y = "round") +
  xlim(1000,9000)

Smooth functions are useful.

coop_max <- coop_rates %>%
    group_by(tournament) %>%
    summarise(max_coop = max(value))

coop_rates %>%
  mutate(seed = as.factor(Seed)) %>%
  ggplot() +
  geom_smooth(aes(round, value, color = seed), se = F) +
  facet_wrap(~tournament) +
  labs(title = "Average cooperation ratio per seed per tournament type.",
       x = "coop ratio",
       y = "round") +
  xlim(1000,9000)

coop_rates %>%
  mutate(seed = as.factor(Seed)) %>%
  ggplot() +
  geom_smooth(aes(round, value), color = "black") +
  facet_wrap(~tournament) +
  labs(title = "Average cooperation ratio per seed per tournament type.",
       x = "coop ratio",
       y = "round") +
  xlim(1000,9000)

outliers %>%
  mutate(seed = as.factor(Seed)) %>%
  ggplot() +
  geom_point(aes(S.D., Counts, color = seed)) +
  geom_smooth(aes(S.D., Counts), color = "black") +
  facet_wrap(~tournament) +
  labs(title = "The instability of the average rage of cooperation.",
       x = "s.d.",
       y = "deviation from average")